Materials and Methods

Cell culture

Both primary cultures of human dermal microvascular endothelial cells (HDMEC), obtained from Emory Skin Diseases Research Center, and HDMEC immortalized with SV40 Large T (5A32 cells) were utilized. Cells were isolated and passaged as described previously (^{Ades et al, 1992;Swerlick et al, 1992a}). Cells were grown in flasks coated with 0.1% gelatin in MCDB 131 media (Mediatech, Herndon, VA) supplemented with 10% (v/v) fetal bovine serum (Hyclone, Logan, Utah), 1% penicillin/streptomycin (Mediatech), 2 mM L-glutamine (Mediatech), 0.25 mg/ml cAMP and 1 g/ml hydrocortisone (Sigma, St. Louis, MO) and maintained at 37°C in a humidified atmosphere of 5% CO₂.

Reagents and antibodies

The iron chelators, 2,2-dipyridyl (DP), a soluble chelator of ferrous iron (Fe²⁺), and desferoxamine (DFO), which binds ferric iron (Fe³⁺), were obtained from Sigma. 5, 5 dimethyl-pyrroline n-oxide (DMPO), a spin trapper without metal binding capability, and 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO), a nitroxide spin trap that can maintain iron as Fe³⁺, were also purchased from Sigma. Recombinant human TNF and IFN were purchased from R & D Systems, Inc. (Minneapolis, MN). Anti-IRF-1 antibody was obtained from Santa Cruz Biotechnology Inc. (Santa Cruz, CA) and rabbit polyclonal antip65 and antip50 antibodies were purchased from Biodesign International (Saco, ME). Anti-phospho-p65 (Ser 536) antibody was purchased from Cell Signaling Technology, Inc (Beverly, MA). The VCAM-1 P8B1 monoclonal antibody developed by E.A. Wayner and T. LeBien was obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by The University of Iowa.

Cell surface ELISA

ELISA using anti-VCAM-1 monoclonal antibodies P3C4 (generous gift of Dr. Elizabeth Wayner) or derived from the VII-6G10 hybridoma (ATCC, Rockville, MD), or using anti-ICAM-1 monoclonal antibody 84H10 (generous gift from Dr. Stephen Shaw) was performed as previously described (^{Ades et al, 1992;Swerlick et al, 1992b}). The cells were treated with DP at doses ranging from 125 to 1000 M for 0–24 h followed by 16 h of TNF. Briefly, the cells were incubated with primary antibodies for one hour, fixed with 2% formaldehyde, and washed with Hanks buffer (with calcium and magnesium). Secondary antibody (goat antimouse HRP conjugate, Bio-Rad, Hercules, CA) was incubated for one hour and then washed thoroughly with Hanks buffer (with calcium and magnesium). Color was developed using DAKO TMB solution for 5–30 min, the reaction stopped, and then the plate read with an ELISA reader (Microplate Autoreader Bio-Tek Instruments EL 311 s) at O.D. 450 nm. Figures represent one representative experiment, and all were repeated at least three times. Bars represent the mean OD of four wells SD.

Flow cytometric analysis

Cell surface expression of VCAM-1 was examined in HDMEC by flow cytometric analysis as previously described (^{Swerlick et al, 1992b}). Briefly, HDMEC were stained for VCAM-1 expression before and after TNF (1000 /ml) for 16 h. Cells were trypsinized, aliquotted, pelleted and stained with anti VCAM-1 monoclonal antibody (P1B8, Hybridoma databank) followed by FITC conjugated goat antimouse IgG antibody (Caltag, Burlingame, CA). Mean channel fluorescence was determined for each data point using a minimal of 10000 events. Cell viability was determined by simultaneous staining with propidium iodide.

Measurement of VCAM-1 mRNA and heterogeneous nuclear VCAM-1 RNA (hnRNA) expression

HDMEC were treated with 125–500 M DP for 24 h followed by 200 U/ml TNF for 24 h. Total RNA was isolated with TriReagent (Sigma) from 5A32 HDMEC according to manufacturer's instructions. cDNA was prepared using the Superscript Preamplification System for 1st Strand cDNA Synthesis (Gibco/BRL, Carlsbad, CA) using the random primer protocol. Real-time polymerase chain reaction (PCR) was done using SYBR green technology with Perkin-Elmer 5700 Gene-Amp Detection System with VCAM-1 primers (forward sequence 5'-CAT GGA ATT CGA ACC CAA ACA-3' and reverse sequence 5'-GGC TGA CCA AGA CGG TTG TATC-3'). Data was normalized with primers for a housekeeping gene. Relative quantities were determined by generating a standard curve from dilutions of cDNA containing the message of interest. Data was then expressed as fold increase over control. Experiments were repeated at least three times, and data shown are a single representative study.

VCAM-1 hnRNA expression was measured to assess the effect of iron chelation on the transcription of the VCAM-1 gene using a modification of the above technique (^{Elferink and Reiners, 1996}). Prior to reverse transcription, total RNA was treated with RNAse-free DNAse for 1 h at room temperature to destroy any potential contaminating genomic DNA. DNAse was inactivated by heating at 75°C for 20 min VCAM-1 primers were selected that would amplify only nascent, unspliced hnRNA (forward sequence 5'-TAC AGT GTT ACT TCT TCT TCC ACA TTC AA-3' located at the intron between exon 6 and 7 and reverse sequence 5'-GCC TCA TGA CTC ACT TTA ACC AATT-3' located at the intron between exon 6 and 7). Emory University Microchemical Facility synthesized all primers.

Electrophoretic mobility-shift assay (EMSA)

Confluent 5A32 HDMEC were treated with 500 U/ml TNF for 2 h after preincubation with 500 M DP for 24 h. Nuclear proteins were extracted and EMSA performed as described previously (^{Gille et al, 1996}). The VCAM-1 oligonucleotide was synthesized to encompass the two NF-kB-like sites of the human VCAM-1 promoter. The double stranded oligonucleotide was 5'-end labeled using the forward reaction with T4 polynucleotide kinase (Gibco/BRL). The DNA binding reaction was performed under the conditions described previously (^{Gille et al, 1996}).

The specificity of NF-kB binding complex was further resolved using competition with a cold unlabeled VCAM-kB oligonucleotide, antibodies recognizing p50, p65 (Biodesign International, Saco, ME), or phosphoserine (Calbiochem, San Diego, CA). Competitor and antibodies were added 20 min before adding the radiolabeled probe to the binding reaction. Samples were subjected to electrophoresis on a native 4% polyacrylamide gel for 4 h at 120 V. Experiments were repeated at least three times, and data shown are a single representative study.

Western blot

Cells were washed with PBS twice and whole cell extracts were prepared by lysing cells in 50 mM, Tris pH 8.0, 150 mM NaCl, 0.02% Na Azide, 1 g/ml Aprotinin, 1% Triton X, 1 mM PMSF. NaF was added for the phosphoserine studies (^{Woetmann et al, 1999}). The suspension was centrifuged at 12,000 rpm for 15 min at 4°C, and the protein in the supernatant was quantified using the BioRad DC Protein Assay. Whole cell protein extract was resolved with SDS-PAGE using a 10% polyacrylamide gel and reducing conditions. After transfer to Hybond ECL nitrocellulose (Amersham Pharmacia Biotech), the protein was stained with Ponceau S (Sigma) to verify uniform loading and transfer. Membranes were blocked with 5% BSA in TBS-T (50 mM Tris-Cl pH 7.4, 150 mM NaCl, 0.1% Tween 20, pH 7.4) overnight and subsequently incubated with primary antibodies (IRF-1 Ab 1:500, anti-p65 Ab 1:2000, anti-phosphoserine 536 p65 Ab 1:1000) for 1h at room temperature. The membrane was washed with TBS-T three times and incubated for 1 h at room temperature with the appropriate horseradish peroxidase-conjugated secondary antibody. Subsequently, the membrane was washed three times with TBS-T and analyzed by enhanced chemiluminescence (ECL Reagent, Amersham Pharmacia Biotech).

Immunofluorescence microscopy

All cells were grown overnight on eight-well chamber slides (Nalge Nunc International, Naperville, IL) before treatment. Designated wells were left untreated or treated with DP for 24 h, DP followed by stimulation with TNF 500 U/ml for 30 min, or TNF alone. The cells were washed with PBS twice for 5 min each and fixed with 3.7% para-formaldehyde for 20 min at room temperature. The cells were washed again and permeabilized with 0.2% Triton X-100 for 10 min at room temperature. The cells were washed and blocked with PBS with 10% FBS at room temperature for at least 10 min. After blocking, the cells were incubated with anti-p65 rabbit antibody (Biodesign) at 1:300 for 1 h at room temperature. The cells were washed three times with blocking solution before adding goat antirabbit FITC conjugated (Caltag, Burlingame, CA) secondary antibody to each well and incubating for 1 h at room temperature in the dark. The cells were washed three times with PBS, mounting reagent added, and the slides visualized with fluorescent microscopy using a Leica DMR-E fluorescence microscope equipped with a Hammamatsu Orca camera. Images were captured and processed using Open Lab software (Improvision Inc., Lexington, MA).

Results

Iron chelators inhibit VCAM-1 induction in HDMEC

When endothelial cells were preincubated with 2,2-dipyridyl (DP), a ferrous iron chelator, and then stimulated with TNF at a dose and for a duration which had been shown previously to maximally induce VCAM-1 expression in HDMEC (^{Ades et al, 1992;Swerlick et al, 1992b}), VCAM-1 protein expression was inhibited in a concentration- and time-dependent manner. Maximal inhibition was observed at doses as low as 500 M (Figure 1a). Partial inhibition was observed when DP was added concurrently with TNF stimulation, with increasing inhibition observed after 8 h (>70%), and complete inhibition requiring at least a 24-h preincubation (Figure 1b). Identical inhibition was observed when expression of induced VCAM-1 was examined by flow cytometric analysis (Figure 1c). No significant decreases in cell viability were noted, even with treatment of cells up to 24 h with doses of DP up to 4 mM when examined by flow cytometric analysis or trypan blue exclusion (data not shown).

Figure 1.

DP pretreatment blocks TNF-mediated VCAM-1 induction in 5A32 HDMEC in a dose and time-dependent manner. HDMEC were pretreated with DP before stimulation with TNF and cell-surface VCAM-1 measured by ELISA. (A) 5A32 HDMEC were pretreated with increasing doses of DP for 8 or 24 h as indicated followed by 500 U/ml TNF for 16 h. (B) 5A32 HDMEC were pretreated with 500 M DP for times ranging from 0 to 24 h and then stimulated with 500 U/ml TNF for 16 h. (C) Flow cytometric analysis of 5A32 HDMEC: Untreated (A), DP 1000 M24 h (B), TNF 1000 U/ml16 h (C), and DP treated (1000 M24 h) followed by TNF (1000 U/ml16 h).

Full figure and legend (23K)

In order to address whether the DP-mediated inhibition of VCAM-1 induction was mediated via iron chelation, we examined the effect of a different iron-specific chelator, desferoxamine (DFO), on TNF-mediated induction of VCAM-1. VCAM-1 protein expression was again inhibited in a concentration-dependent fashion (Figure 2a) with 50% inhibition of TNF-induced VCAM-1 expression seen at 500 M, and maximal inhibition noted at 2000 M. DFO is less cell-permeable than DP, which probably accounts for the need for higher doses for comparable effects. When exogenous iron in the form of ferric citrate (2000 M) was combined with DFO prior to preincubation, inhibition of TNF-mediated VCAM-1 induction was partially reversed (Figure 2b). In addition, when 5A32 HDMEC were treated with TEMPO, a nitroxide spin trap that maintains iron as Fe³⁺ (^{Voest et al, 1993}), and then stimulated with TNF, VCAM-1 induction was also inhibited in a concentration-dependent fashion (Figure 2d). However, DMPO, a spin trap without metal binding capability (^{Rosen and Rauckman, 1984}) did not inhibit TNF induced VCAM-1 up-regulation (Figure 2c). These results further support a role for iron in expression of VCAM-1.

Figure 2.

Iron chelation activity is required for VCAM-1 blockade. (A) 5A32 cells were pretreated with increasing concentrations of the iron chelator DFO for 4 h prior to stimulation with TNF (100 U/ml16 h). Cell surface VCAM-1 expression was measured by ELISA. (B) DFO- mediated inhibition of VCAM-1 induction was partially reversed when cells were pretreated with 200 M DFO plus ferric citrate (2000 M4 h) and stimulated with TNF (500 U/ml16 h). 5A32 HDMEC were pretreated with the indicated concentrations of DMPO (C) and TEMPO (D) for 2 h and were stimulated with TNF (500 U/ml16 h). TNF-mediated VCAM-1 expression was blocked in a dose-dependent manner by TEMPO, a metal binding spin trapper that can maintain iron as Fe³⁺, but not by DMPO, a nonmetal binding spin trapper, providing further support for a role of iron in VCAM-1 expression in 5A32 HDMEC.

Full figure and legend (41K)

Inhibition of VCAM-1 induction is not cell- or gene-specific

To determine whether the DP mediated inhibition was gene or cell-specific, we examined the effect of DP pretreatment on TNF mediated VCAM-1 induction in human umbilical vein endothelial cells (HUVEC) and on TNF mediated ICAM-1 induction in HDMEC. DP inhibited VCAM-1 induction of VCAM-1 in HUVEC in a dose dependent manner virtually identical to that observed in HDMEC. Unlike the case for VCAM-1 expression, 8 h of pretreatment with DP (500 M) had virtually no effect on ICAM-1 cell surface expression. However, preincubation of cells with DP for 24 h resulted in marked inhibition of ICAM-1 protein expression (Figure 3).

Figure 3.

DP mediated inhibition is not cell- or gene specific: (A) HUVEC were pretreated with DP (30–1000 M) prior to stimulation with TNF (1000 U/ml16 h). Cells were assayed for VCAM-1 expression by cell based ELISA. (B) 5A32 HDMEC were pretreated with various concentrations of DP for 8 or 24 h followed by 500 U/ml TNF for 16 h. Cell-surface levels of ICAM-1 were measured by ELISA. Preincubation of cells with DP for 24 h prior to cytokine treatment significantly blocked TNF-mediated ICAM-1 induction.

Full figure and legend (28K)

Iron chelators inhibit VCAM-1 gene transcription

In order to examine whether inhibition of VCAM-1 expression may be mediated by changes in gene expression, we initially examined whether DP pretreatment prior to TNF stimulation inhibited TNF-mediated increases in VCAM-1 mRNA expression. DP pretreatment inhibited up-regulation of VCAM-1 mRNA by >90% (Figure 4a). Similarly, DFO almost completely inhibited TNF-mediated increases in VCAM-1 mRNA levels (Figure 4b). In contrast, the nonmetal binding spin trap DMPO had no effect, consistent with its lack of effect on TNF induced VCAM-1 protein expression (Figure 4b).

Figure 4.

DP and DFO but not DMPO inhibits TNF-mediated VCAM-1 mRNA up-regulation. (A) 5A32 HDMEC were pretreated with 250 or 125 M DP for 24 h and then stimulated with 200 U/ml TNF for 24 h. VCAM-1 mRNA was measured by quantitative RT-PCR and differences represented as fold increase over baseline. (B) Confluent monolayers of 5A32 HDMEC were pretreated with 500 M DP, 500 M DFO, or 500 M DMPO for 24 h and then stimulated with 200 U/ml TNF for 24 h. DP and DFO inhibit TNF-mediated increases in VCAM-1 mRNA, but DMPO did not.

Full figure and legend (27K)

In order to directly examine whether iron chelators inhibit VCAM-1 expression by inhibition of VCAM-1 gene transcription, we utilized real time PCR to measure the expression of VCAM-1 hnRNA after cytokine stimulation in the presence and absence of the iron chelator DP. PCR primer pairs were designed that recognized intronic sequences that are found only in unspliced VCAM-1 hnRNA. Treatment of cells with TNF resulted in a marked induction of hnRNA that was inhibited by pretreatment with actinomycin D or DP (Figure 5). These data provide support for VCAM-1 gene transcription as being the target for iron chelator mediated inhibition of VCAM-1 expression.

Figure 5.

DP inhibits TNF-mediated VCAM-1 expression by inhibition of VCAM-1 gene transcription. 5A32 HDMEC were pretreated with 500 M DP for 8 h or 2 g/ml actinomycin D for 30 min prior to stimulation with 500 U/ml TNF for 4 h. The fold differences of VCAM-1 hnRNA were determined by quantitative RT-PCR. Results are representative of three experiments.

Full figure and legend (9K)

Iron chelators do not inhibit NF-kB translocation or DNA binding

Iron chelators have been shown to regulate the stability of transcription factors (^{Wang and Semenza, 1993;Lin et al, 1997}). Previous studies have also demonstrated that certain antioxidants may also block translocation of activated NF-kB complexes from the cytoplasm to the nucleus, thus inhibiting NF-kB mediated gene activation (^{Bowie et al, 1997}). We hypothesized that iron chelator-mediated inhibition of VCAM-1 induction may be controlled via effects on NF-kB stability or function. In order to test this hypothesis, we examined nuclear localization of NF-kB complexes after TNF stimulation using immunofluorescence microscopy. TNF stimulation of HDMEC resulted in prompt translocation of p65 from the cytoplasm to the nucleus. However, DP pretreatment prior to TNF had no detectable effect on this TNF-mediated p65 translocation (Figure 6).

Figure 6.

DP does not affect NF-kB translocation after TNFstimulation in 5A32 HDMEC cells. (A) Untreated cells; or (B) treated with 500 U/ml TNF for 30 min; (C) 500 M DP for 24 h; and (D) 500 U/ml TNF for 30 min after pretreatment with DP 500 M for 24 h. Cellular localization of p65 was identified by immunofluorescence microscopy.

Full figure and legend (116K)

It was possible that although DP pretreatment may not affect NF-kB translocation, DP may alter the ability of translocated complexes to bind to relevant sequences in the VCAM-1 promoter. In order to examine whether DP pretreatment resulted in loss of NF-kB binding activity despite apparently normal translocation, nuclear extracts were examined for protein binding to the VCAM-1 NF-kB sequence using EMSA (Figure 7). DP treatment had no effect on the binding of translocated NF-kB complexes. After TNF stimulation (lanes 3–8), complex formation was the same in both DP pretreated (lanes 6–8) and untreated samples (lanes 3–5). New complex formation (bands A and B, lanes 3 and 6), which was shifted with addition of antip50 (band C, lanes 4 and 7) or antip65 (band D, lanes 5 and 8) antibodies was also identical in each case. These data would suggest that the DP effect is mediated via targets other than NF-kB translocation or DNA binding.

Figure 7.

DP does not affect NF-kB binding to the VCAM-1 kB element after TNF stimulation. 5A32 HDMEC were stimulated with 500 U/ml TNF for 2 h with or without pretreatment with DP (500 M) for 24 h. Nuclear extracts were assessed for the binding of NF-kB by EMSA. TNF treatment results in appearance of complexes A and B, which appear in cells regardless of whether pretreated with DP. In addition, both bands are supershifted in treated cells by antip50 antibody (band C) or antip65 antibody (band D) with or without DP pretreatment.

Full figure and legend (133K)

Iron chelators inhibit TNF-mediated IRF-1 induction

The previous data suggested that the effect of DP on VCAM-1 gene transcription may be mediated via an NF-kB independent mechanism. Another potential target is interferon regulatory factor 1 (IRF-1), a transcription factor that is also induced by TNF. In addition to the tandem NF-kB elements in the VCAM-1 promoter, optimal transcriptional induction of VCAM-1 depends to some degree upon binding of IRF-1 to a distinct and adjacent element in the VCAM-1 promoter (^{Neish et al, 1995;Lechleitner et al, 1998}). In order to examine whether iron chelation blocks IRF-1 expression, we examined IRF-1 protein expression after cytokine stimulation with and without DP pretreatment. Treatment of HDMEC with either TNF or IFN induced expression of IRF-1 protein. Pretreatment of 5A32 HDMEC with DP prior to either TNF or IFN treatment completely inhibited the expression of IRF-1 protein expression. This inhibition was dependent upon DP concentration and preincubation time, roughly paralleling DP effects on VCAM-1 induction. Thus, DP may inhibit VCAM-1 induction via inhibition of IRF-1 induction (Figure 8a).

Figure 8.

Pretreatment of 5A32 HDMEC cells with IFN prior to DP pretreatment induces persistent IRF-1 expression but only partially restores TNF-mediated VCAM-1 induction. (A) 5A32 HDMEC were pretreated for 4, 8 and 24 h with 1000 M DP prior to TNF stimulation. Cell lysates were assayed for the presence of IRF-1 by Western blotting. Untreated cells (lane 1), DP treated (500 M24 h, lane 2), TNF treated (500 U/ml2 h, lane 3) or 2 h of TNF treatment preceded by incubations of DP (500 M) for 4 h (lane 4), 8 h (lane 5), and 24 h (lane 6). (B) IFN induces IRF-1 expression that persists after DP treatment. Untreated cells (lane 1), IFN (1000 U/ml) for 2 h (lane 2), IFN (1000 U/ml) for 10 h (lane 3), IFN (1000 U/ml)2 h followed by DP (1000 M) for 8 h (lane 4), and IFN (1000 U/ml) for 2 h followed by DP (1000 M) for 24 h (lane 5). Cell lysates were prepared and assessed for the presence of IRF-1 by Western blotting. (C) Prior induction of IRF-1 by IFN does not restore VCAM-1 induction inhibited by DP pretreatment as measured by ELISA. Control cells (untreated), IFN (1000 U/ml 24 h), DP (500 M24 h), IFN/DP (IFN 1000 U/ml2 h followed by DP 500 M24 h), TNF 500 U/ml16 h), IFN/TNF (IFN 1000 U/ml2 h then TNF 500 U/ml16 h), DP/TNF (DP 200 M6 h then TNF 500 U/ml16 h), and IFN/DP/TNF (IFN 1000 U/ml2 h, then DP 500 M6 h, then TNF 500 U/ml16 h). Pretreatment of the cells with IFN and induction of IRF-1 expression prior to DP pretreatment only partially restores TNF-mediated VCAM-1 induction.

Full figure and legend (86K)

In order to examine whether the DP mediated inhibition of VCAM-1 induction was primarily mediated via effects on IRF-1 expression, we preinduced IRF-1 expression in 5A32 cells with IFN. By doing so, we were able to induce persistent IRF-1 expression without inducing VCAM-1 and IRF-1 expression persisted for up 8 h, even after the addition of DP. HDMEC were then treated with TNF, with and without DP pretreatment (Figure 8b). We then utilized cell based ELISA to determine whether preinduction of IRF-1 by IFN could restore VCAM-1 induction by TNF. However, induction of IRF-1 with IFN prior to DP treatment only partially restored TNF-mediated VCAM-1 induction (Figure 8c). These data suggest that although inhibition of IRF-1 expression by DP may play a partial role in DP mediated-VCAM-1 inhibition additional targets for DP effect needed to be defined.

Iron chelators do not inhibit NF-kB phosphorylation

Previous studies have suggested that changes in NF-kB phosphorylation, particularly at serine residues, may be critical for the ability of NF-kB proteins, particularly p65, to transactivate genes without affecting DNA binding (^{Wang and Baldwin, 1998;Wang et al, 2000}). To examine whether DP pretreatment inhibited p65 phosphorylation, we utilized antibody in western analysis that specifically recognized serine-phosphorylated p65. Under basal conditions, some p65 was constitutively phosphorylated and TNF stimulation resulted in a prompt increase in the level of phosphorylation. Pretreatment of HDMEC with DP had no effect on constitutive phosphorylation or the increases associated with TNF treatment (Figure 9).

Figure 9.

DP does not globally inhibit serine phosphorylation of p65. 5A32 HDMEC cells were pretreated for 20 h with 500 M of DP prior to stimulation with 500 U/ml TNF for 30 min Cell lysates (A) untreated HDMEC (B) TNF 500 U/ml30 min (C) DP 500 M24 h (D) DP 500 M24 h, then TNF 500 U/ml30 min, were assessed for the presence of phospho-Ser 536 p65 antibody (Cell Signaling Technology) by Western blotting. The same blots were then stripped and probed with antip65 antibody.

Full figure and legend (24K)

Iron chelators inhibit cytokine induction of NF-kB genes with delayed kinetics, but not IkB

Despite blocking the TNF mediated induction of VCAM-1 and ICAM-1, DP appeared to have not discernable effect on NF-kB translocation, binding, or phosphorylation that might affect its transactivating function. Alternative targets requiring iron are mechanisms regulating chromatin structure and availability of promoters of NF-kB responsive genes. Previous studies have shown that genes that are rapidly activated after cytokine stimulation, such as IkB, do not require chromatin modification for transcription factor accessibility (^{Saccani et al, 2001;Saccani et al, 2002}). In order to examine whether DP may affect VCAM-1 expression via this target, we examined the TNF induction of both genes with constitutively and immediately accessible (CIA) promoters (IkB) and those with delayed inductions that presumably require stimulus-dependent modifications in chromatin structure to make NF-kB sites accessible (^{Saccani et al, 2001;Saccani et al, 2002}). DP pretreatment blocked induction of mRNA of two additional NF-kB regulated genes, interleukin-8 (Figure 10) and tissue factor (data not shown), in HDMEC. In contrast, DP had no effect on the TNF mediated up-regulation of IkB mRNA (Figure 10). These data are consistent with DP targeting cytokine-mediated effects on chromatin packaging.

Figure 10.

DP blocks TNF mediated increases in IL-8 mRNA but does not inhibit increases in IkB mRNA: 5A32 HDMEC were stimulated with TNF (1000 U/ml4 h) with and without pretreatment with DP (1000 M24 h). Steady state levels of IL-8 and IkB mRNA were assessed by real-time quantitative PCR as described in material and methods. Levels are expressed as fold-increase over unstimulated, untreated controls.

Full figure and legend (16K)

Discussion

Iron has been implicated as a key participant in a variety of inflammatory disorders and processes. Studies examining the effect of iron chelators on the function of key transcription factors such as NF-kB and HIF-1 have shown that chelation of iron affects proteasomal processing, resulting in inhibited degradation (^{Wang and Semenza, 1993;Bowie et al, 1997}). Given the dependence of VCAM-1 expression on proteasomal degradation of the IkB regulatory subunit of NF-kB, it appeared to be reasonable to predict that iron chelators might inhibit VCAM-1 induction by TNF. Consistent with our hypothesis, the iron chelator DP almost completely inhibited TNF-mediated induction of VCAM-1 at the protein and mRNA level.

Similar effects upon TNF-mediated induction of VCAM-1 were also observed in HDMEC with pretreatment using a different iron specific chelator, DFO. Addition of exogenous iron in the form of ferric citrate to DFO before stimulation with TNF partially reversed the chelator effect, supporting again iron binding as the target of action of iron chelators. The electron spin trap 5-dimethyl-1-pyrrolone-N-oxide (DMPO), which does not bind to iron, failed to inhibit TNF mediated induction of VCAM-1. However, when HDMEC were treated with TEMPO, a nitroxide spin trap that maintains iron as Fe³⁺, and then stimulated with TNF, VCAM-1 induction was inhibited in a dose-dependent fashion. These data further support the role of iron in TNF-mediated VCAM-1 up-regulation.

Our data clearly point to VCAM-1 gene transcription as a target for DP. DP pretreatment completely blocked TNF mediated increases in VCAM-1 steady state mRNA expression, and hnRNA induction. However, the exact mechanism of this effect is not completely defined. Iron is required for activity of an intracellular prolyl hydroxylase that modifies HIF-1, facilitating its interaction with Von Hippel Lindau protein and proteasomal degradation (^{Ivan et al, 2001}). Similarly, DFO has also been shown to inhibit TNF mediated IkB degradation. We hypothesized that iron chelation may inhibit VCAM-1 gene activation by blocking TNF-mediated IkB degradation and NF-kB activation. However, iron chelators had no effect on either NF-kB translocation or the ability of translocated complexes to bind to VCAM-1 kB elements, even at doses sufficient to block VCAM-1 protein and mRNA expression. Given the lack of obvious effect of DP pretreatment on NF-kB binding to the VCAM-1 kB element, other potential targets for iron chelators had to be considered.

VCAM-1 expression has also been shown to be sensitive to antioxidants such as n-acetylcysteine (NAC) and pyrrolidinedithiocarbamate (PDTC) (^{Marui et al, 1993}). Agents such as PDTC are also metal chelators, and it has been proposed that their potent effects may be mediated via a combination of metal chelation and related antioxidant effects (^{Saran and Bors, 1990}). It is possible that chelation of iron results in loss of metal dependent redox signaling. However, our data does not support this explanation since antioxidant mediated inhibition of VCAM-1 induction has been shown to be mediated via inhibition of NF-kB activation and translocation, and DP had no effect whatsoever on NF-kB activation, translocation, or DNA binding (^{Bowie et al, 1997}).

Alternatively, iron chelators may inhibit VCAM-1 induction by blocking the induction of IRF-1 expression. Previous studies have suggested that although IRF-1 augments the expression of VCAM-1 transcription, it is not necessary for gene expression (^{Neish et al, 1995;Lechleitner et al, 1998}). Our data are consistent with these previous observations. Pretreatment of endothelial cells with DP completely blocked TNF-mediated induction of IRF-1 expression. However, prior induction of IRF-1 by treatment with IFN only partially restored TNF-mediated VCAM-1 induction. Furthermore, DP pretreatment blocks TNF mediated up-regulation of ICAM-1 expression in HDMEC, a process that is NF-kB dependent and IRF-1 independent (^{Gille et al, 1997;Paxton et al, 1997}).

It is possible that iron is required for NF-kB complexes to transactivate the VCAM-1 gene without affecting binding to sites in the promoter as assessed by EMSA. Previous studies have shown that anti-inflammatory agents such as sulfasalazine inhibit the p65 phosphorylation, and this phosphorylation is critical for p65-mediated gene transactivation (^{Egan et al, 1999}). The target for phosphorylation may be one or more of a series of serine residues. Multiple enzymes, which induce phosphorylation of specific serine residues, have been described (^{Wang and Baldwin, 1998;Wang et al, 2000}). Our EMSA studies do not demonstrate any global inhibition of NF-kB serine phosphorylation by DP. In addition, iron chelators do not block TNF-mediated phosphorylation of Ser536, a residue previously identified for its role in p65 function (^{Sakurai et al, 1999}). However, our studies cannot rule out an effect on phosphorylation of other residues such as Ser529 (^{Wang and Baldwin, 1998;Wang et al, 2000}). Iron may still be critical for the phosphorylation of specific serine residues in p65, and altered patterns of phosphorylation may affect VCAM-1 gene transcription without affecting translocation or DNA binding.

Paradoxically, although DP did not affect the activation of NF-kB or its ability to bind in vitro to oligonucleotides corresponding to VCAM-1 kB elements, it still almost completely blocked VCAM-1 gene transcription mediated by TNF. It is possible that iron blocked VCAM-1 induction in HDMEC via a mechanism completely independent of NF-kB. This is quite unlikely given previous studies of TNF mediated VCAM-1 gene expression in this context showing a strict dependence on NF-kB (^{Iademarco et al, 1992;Neish et al, 1995;Gille et al, 1996}). An alternative mechanism may be an iron dependent process regulating chromatin modifications and accessibility. The effects of iron chelators on TNF mediated genes in HDMEC are consistent with an effect on DNA packaging and histone modification. The differential effects on VCAM-1 and IkB could be explained on the basis of differential requirements for chromatin remodeling as a prerequisite for transcription (^{Saccani et al, 2001}). Certain NF-kB driven genes are not immediately accessible and require modification of histones in addition to NF-kB translocation for expression. The transcription of these genes tends to be delayed after cytokine stimulation and NF-kB activation, presumably because histone modification requires additional time. In contrast, other NF-kB driven genes such as IkB are immediately accessible and are rapidly induced after cytokine stimulation. We reasoned that if iron was important for chromatin accessibility, iron chelators should not have an effect on genes such as IkB that were constitutively accessible while iron chelators should inhibit expression of NF-kB inducible genes with delayed kinetics. Our observations that DP pretreatment had no effect on IkB induction, while consistently blocking other NF-kB inducible genes with delayed kinetics, support this hypothesis.

Previous studies linking chromatin remodeling to iron also provide additional indirect evidence for the role of iron. Iron is essential for expression of a number of genes critical for cell growth and proliferation, and removal of iron via the use of DFO has effects that are similar to the effects of histone deacetylase (HDAC) (^{Kim et al, 2001}) (^{Kramer et al, 2002}). Treatment of cells with DFO resulted in cell cycle arrest and down-regulation of p21^WAF and specific cyclin dependent kinases. From these data, one can conclude that expression of p21^WAF is iron dependent (^{Kramer et al, 2002}). Similarly, HDAC inhibition resulted in gene expression comparable to iron loading, with resultant expression of p21^WAF and cell proliferation (^{Kim et al, 2001}). Consistent with our hypothesis is the converse of these observations. Histone deacetylation has effects that are essentially identical to iron chelation. It is reasonable to propose that they may operate via a common mechanism and that mechanism likely involves chromatin accessibility.

In conclusion, iron chelators inhibited VCAM-1 protein and mRNA induction in a concentration- and time-dependent manner via a mechanism that appears to be transcriptionally mediated. The target of iron chelator effects was not NF-kB activation, but appeared to be mediated through a combination of inhibition of IRF-1 expression and effects on the ability of translocated NF-kB complexes to transactivate the VCAM-1 gene, potentially via iron dependent alterations in chromatin remodeling.

References

1. Ades, EW, Candal, FJ, Swerlick, RA, George, VG, Summers, S, Bosse, DC, Lawley, TJ: HMEC-1: Establishment of an immortalized human microvascular endothelial cell line. J Invest Dermatol 1992 99: 683–690,  | Article | PubMed | ISI | ChemPort |
2. Allen, RG, Tresini, M: Oxidative stress and gene regulation. Free Radic Biol Med 2000 28: 463–499,  | Article | PubMed | ISI | ChemPort |
3. Baeuerle, PA, Henkel, T: Function and activation of NF-kappa B in the immune system. Annu Rev Immunol 1994 12: 141–179,  | Article | PubMed | ISI | ChemPort |
4. Bowie, AG, Moynagh, PN, O'Neill, LA: Lipid peroxidation is involved in the activation of NF-kappaB by tumor necrosis factor but not interleukin-1 in the human endothelial cell line ECV304: Lack of involvement of H2O2 in NF-kappaB activation by either cytokine in both primary and transformed endothelial cells. J Biol Chem 1997 272: 25941–25950,  | Article | PubMed | ISI | ChemPort |
5. Brenneisen, P, Wenk, J, Klotz, LO, et al: Central role of Ferrous/Ferric iron in the ultraviolet B irradiation-mediated signaling pathway leading to increased interstitial collagenase (matrix-degrading metalloprotease (MMP)-1) and stromelysin-1 (MMP-3) mRNA levels in cultured human dermal fibroblasts. J Biol Chem 1998 273: 5279–5287,  | Article | PubMed | ISI | ChemPort |
6. Das, PK, de Boer, OJ, Visser, A, Verhagen, CE, Bos, JD, Pals, ST: Differential expression of ICAM-1, E-selectin and VCAM-1 by endothelial cells in psoriasis and contact dermatitis. Acta Derm Venereol 1994 186(Suppl)21–22,  | ChemPort |
7. Egan, LJ, Mays, DC, Huntoon, CJ, et al: Inhibition of interleukin-1-stimulated NF-kappaB RelA/p65 phosphorylation by mesalamine is accompanied by decreased transcriptional activity. J Biol Chem 1999 274: 26448–26453,  | Article | PubMed | ISI | ChemPort |
8. Elferink, CJ, Reiners, JJ Jr: Quantitative RT-PCR on CYP1A1 heterogeneous nuclear RNA: A surrogate for the in vitro transcription run-on assay. Biotechniques 1996 20: 470–477,  | PubMed | ISI | ChemPort |
9. Elices, MJ, Osborn, L, Takada, Y, Crouse, C, Luhowskyj, S, Hemler, ME, Lobb, RR: VCAM-1 on activated endothelium interacts with the leukocyte integrin VLA-4 at a site distinct from the VLA-4/fibronectin binding site. Cell 1990 60: 577–584,  | Article | PubMed | ISI | ChemPort |
10. Gille, J, Paxton, LL, Lawley, TJ, Caughman, SW, Swerlick, RA: Retinoic acid inhibits the regulated expression of vascular cell adhesion molecule-1 by cultured dermal microvascular endothelial cells. J Clin Invest 1997 99: 492–500,  | Article | PubMed | ISI | ChemPort |
11. Gille, J, Swerlick, RA, Lawley, TJ, Caughman, SW: Differential regulation of vascular cell adhesion molecule-1 gene transcription by tumor necrosis factor alpha and interleukin-1 alpha in dermal microvascular endothelial cells. Blood 1996 87: 211–217,  | PubMed | ISI | ChemPort |
12. Groves, RW, Ross, EL, Barker, JN, MacDonald, DM: Vascular cell adhesion molecule-1: Expression in normal and diseased skin and regulation in vivo by interferon gamma. J Am Acad Dermatol 1993 29: 67–72,  | PubMed | ISI | ChemPort |
13. Iademarco, MF, McQuillan, JJ, Rosen, GD, Dean, DC: Characterization of the promoter for vascular cell adhesion molecule-1 (VCAM-1). J Biol Chem 1992 267: 16323–16329,  | PubMed | ISI | ChemPort |
14. Ivan, M, et al: HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: Implications for O2 sensing. Science 2001 292: 464–468,  | Article | PubMed | ISI | ChemPort |
15. Kim, JS, Lee, S, Lee, T, Lee, YW, Trepel, JB: Transcriptional activation of p21 (WAF1/CIP1) by apicidin, a novel histone deacetylase inhibitor. Biochem Biophys Res Commun 2001 281: 866–871,  | Article | PubMed | ISI | ChemPort |
16. Kramer, JL, Baltathakis, I, Alcantara, OS, Boldt, DH: Differentiation of functional dendritic cells and macrophages from human peripheral blood monocyte precursors is dependent on expression of p21 (WAF1/CIP1) and requires iron. Br J Haematol 2002 117: 727–734,  | Article | PubMed | ISI | ChemPort |
17. Lechleitner, S, Gille, J, Johnson, DR, Petzelbauer, P: Interferon enhances tumor necrosis factor-induced vascular cell adhesion molecule 1 (CD106) expression in human endothelial cells by an interferon-related factor 1-dependent pathway. J Exp Med 1998 187: 2023–2030,  | Article | PubMed | ISI | ChemPort |
18. Lin, M, Rippe, RA, Niemela, O, Brittenham, G, Tsukamoto, H: Role of iron in NF-kappa B activation and cytokine gene expression by rat hepatic macrophages. Am J Physiol 1997 272: G1355–G1364,  | PubMed | ISI | ChemPort |
19. Litchfield, TM, Smith, CH, Atkinson, BA, Norris, PG, Elliott, P, Haskard, DO, Lee, TH: Eosinophil infiltration into human skin is antigen-dependent in the late-phase reaction. Br J Dermatol 1996 134: 997–1004,  | Article | PubMed | ISI | ChemPort |
20. Lontz, W, Sirsjo, A, Liu, W, Lindberg, M, Rollman, O, Torma, H: Increased mRNA expression of manganese superoxide dismutase in psoriasis skin lesions and in cultured human keratinocytes exposed to IL-1 beta and TNF-alpha. Free Radic Biol Med 1995 18: 349–355,  | Article | PubMed | ISI | ChemPort |
21. Marui, N, Offermann, MK, Swerlick, R, et al: Vascular cell adhesion molecule-1 (VCAM-1) gene transcription and expression are regulated through an antioxidant-sensitive mechanism in human vascular endothelial cells. J Clin Invest 1993 92: 1866–74,  | Article | PubMed | ISI | ChemPort |
22. McCord, JM: Iron, free radicals, and oxidative injury. Semin Hematol 1998 35: 5–12,  | PubMed | ISI | ChemPort |
23. Neish, AS, Read, MA, Thanos, D, Pine, R, Maniatis, T, Collins, T: Endothelial interferon regulatory factor 1 cooperates with NF-kappa B as a transcriptional activator of vascular cell adhesion molecule 1. Mol Cell Biol 1995 15: 2558–2569,  | PubMed | ISI | ChemPort |
24. Norris, P, Poston, RN, Thomas, DS, Thornhill, M, Hawk, J, Haskard, DO: The expression of endothelial leukocyte adhesion molecule-1 (ELAM-1), intercellular adhesion molecule-1 (ICAM-1), and vascular cell adhesion molecule-1 (VCAM-1) in experimental cutaneous inflammation: A comparison of ultraviolet B erythema and delayed hypersensitivity. J Invest Dermatol 1991 96: 763–770,  | Article | PubMed | ISI | ChemPort |
25. Norton, J, Sloane, JP, al-Saffar, N, Haskard, DO: Vessel associated adhesion molecules in normal skin and acute graft-versus-host disease. J Clin Pathol 1991 44: 586–591,  | PubMed | ISI | ChemPort |
26. Onuma, S: Immunohistochemical studies of infiltrating cells in early and chronic lesions of psoriasis. J Dermatol 1994 21: 223–232,  | PubMed | ChemPort |
27. Paxton, LL, Li, LJ, Secor, V, Duff, JL, Naik, SM, Shibagaki, N, Caughman, SW: Flanking sequences for the human intercellular adhesion molecule-1 NF-kappaB response element are necessary for tumor necrosis factor alpha-induced gene expression. J Biol Chem 1997 272: 15928–15935,  | Article | PubMed | ISI | ChemPort |
28. Pourzand, C, Watkin, RD, Brown, JE, Tyrrell, RM: Ultraviolet A radiation induces immediate release of iron in human primary skin fibroblasts: The role of ferritin. Proc Natl Acad Sci USA 1999 96: 6751–6756,  | Article | PubMed | ChemPort |
29. Rosen, GM, Rauckman, EJ: Spin trapping of superoxide and hydroxyl radicals. Meth Enzymol 1984 105: 198–209,  | PubMed | ISI | ChemPort |
30. Saccani, S, Pantano, S, Natoli, G: p38-Dependent marking of inflammatory genes for increased NF-kappa B recruitment. Nat Immunol 2002 3: 69–75,  | Article | PubMed | ISI | ChemPort |
31. Saccani, S, Pantano, S, Natoli, G: Two waves of nuclear factor kappa B: Recruitment to target promoters. J Exp Med 2001 193: 1351–1359,  | Article | PubMed | ISI | ChemPort |
32. Sakurai, H, Chiba, H, Miyoshi, H, Sugita, T, Toriumi, W: IkappaB kinases phosphorylate NF-kappaB p65 subunit on serine 536 in the transactivation domain. J Biol Chem 1999 274: 30353–30356,  | Article | PubMed | ISI | ChemPort |
33. Saran, M, Bors, W: Radical reactions in vivo: An overview. Radiat Environ Biophys 1990 29: 249–262,  | Article | PubMed | ISI | ChemPort |
34. Schreck, R, Rieber, P, Baeuerle, PA: Reactive oxygen intermediates as apparently widely used messengers in the activation of the NF-kappa B transcription factor and HIV-1. Embo J 1991 10: 2247–2258,  | PubMed | ISI | ChemPort |
35. Sharkey, P, Eedy, DJ, Burrows, D, McCaigue, MD, Bell, AL: A possible role for superoxide production in the pathogenesis of contact dermatitis. Acta Derm Venereol 1991 71: 156–159,  | PubMed | ISI | ChemPort |
36. Silber, A, Newman, W, Reimann, KA, Hendricks, E, Walsh, D, Ringler, DJ: Kinetic expression of endothelial adhesion molecules and relationship to leukocyte recruitment in two cutaneous models of inflammation. Lab Invest 1994 70: 163–175,  | PubMed | ISI | ChemPort |
37. Swerlick, RA, Lee, KH, Li, LJ, Sepp, NT, Caughman, SW, Lawley, TJ: Regulation of vascular cell adhesion molecule 1 on human dermal microvascular endothelial cells. J Immunol 1992b 149: 698–705,  | PubMed | ISI | ChemPort |
38. Swerlick, RA, Lee, KH, Wick, TM, Lawley, TJ: Human dermal microvascular endothelial but not human umbilical vein endothelial cells express CD36 in vivo and in vitro. J Immunol 1992a 148: 78–83,  | PubMed | ISI | ChemPort |
39. Voest, EE, van Faassen, E, Marx, JJ: An electron paramagnetic resonance study of the antioxidant properties of the nitroxide free radical TEMPO. Free Radic Biol Med 1993 15: 589–595,  | Article | PubMed | ISI | ChemPort |
40. Wakita, H, Sakamoto, T, Tokura, Y, Takigawa, M: E-selectin and vascular cell adhesion molecule-1 as critical adhesion molecules for infiltration of T lymphocytes and eosinophils in atopic dermatitis. J Cutan Pathol 1994 21: 33–39,  | Article | PubMed | ISI | ChemPort |
41. Wang, D, Baldwin, AS Jr: Activation of nuclear factor-kappaB-dependent transcription by tumor necrosis factor-alpha is mediated through phosphorylation of RelA/p65 on serine 529. J Biol Chem 1998 273: 29411–29416,  | Article | PubMed | ISI | ChemPort |
42. Wang, GL, Semenza, GL: Desferrioxamine induces erythropoietin gene expression and hypoxia-inducible factor 1 DNA-binding activity: Implications for models of hypoxia signal transduction. Blood 1993 82: 3610–3615,  | PubMed | ISI | ChemPort |
43. Wang, D, Westerheide, SD, Hanson, JL, Baldwin, AS Jr: Tumor necrosis factor alpha-induced phosphorylation of RelA/p65 on Ser529 is controlled by casein kinase II. J Biol Chem 2000 275: 32592–32597,  | Article | PubMed | ISI | ChemPort |
44. Woetmann, A, Nielsen, M, Christensen, ST, et al: Inhibition of protein phosphatase 2A induces serine/threonine phosphorylation, subcellular redistribution, and functional inhibition of STAT3. Proc Natl Acad Sci USA 1999 96: 10620–10625,  | Article | PubMed | ChemPort |

Acknowledgments

This work was supported by Post-doctoral Fellowship Program of Korea Science and Engineering Foundation (KOSEF) in 1998, NIH R01 AR39632, Emory Skin Disease Research Center (NIH P30 AR42687). We are grateful for the technical assistance of Mrs. Patricia Kowalczyk and Mr. Marvin Newton-West for their technical assistance.